* UPDATE:
The original article below was from August 2010, when World of Warcraft's total subscriber base was 11.5 million. However, for a more updated version (e.g. 12 million subscribers and taking a different approach to estimate numbers), please look here.
____________________________________________________
Unfortunately, mmogchart.com stopped tracking World of Warcraft's growth back in 2007. However, based on announcements, Blizzard has stated that they have over 11.5 million "Active Subscribers".
Here's an approximate chart of their growth:
Actual Sales Figures
Something always troubled me though..
According to vgchartz, here are Blizzard's most recent sales figures:
How is it possible that they have 11.5M "Active Subscribers" but less than 7M upgraded to WOTLK (released almost 2 years ago) and 7.4M upgraded to Burning Crusade? Let alone, 4.5M players are still playing Vanilla WoW, and there are over 11M players that have been apparently active for over 6 years. Something funny is definitely going on. =]
In February 2010, CEO Mike Morhaime stated that only 30 percent of new World of Warcraft players ever make it past level 10. That's an extremely low retention rate. He also stated that subscription levels had not grown since December 2008 - they had the same number of subscribers now that they did then.
Given their sales numbers, low retention rate, but high claims of "active subscribers" - it just seems impossible. So, let's look at the factual numbers.
Investor Statements
Here are their financials over the most recent 3 month period:
World of Warcraft Gross Profits: $301.75 million
World of Warcraft Operating Costs: $12.75 million (Source)
World of Warcraft GAAP Net Revenue: $289 million (Latest Quarterly Report)
(Note: GAAP means "Generally Accepted Accounting Principles" but when corporations use non-GAAP methodologies, they can inflate or stretch numbers in their books. So it's safe to say that GAAP is the more "honest" figure. If you review their other Quarterly Reports for 2008 and 2009, they also show similar rates.)
Sales during those months were negligible, so their monthly subscriptions totaled approx. $301.75 million gross worldwide.
At a rate of $15 per month, that only accounts for 6.7 million active subscribers! That's almost 50 percent!
(Note: Chinese players pay on average ~$7.27 USD per month and account for ~23% of all worldwide players. However, most CWOW players have multiple accounts. All other geographies pay approx. $15 per month. More details here.)
But they said 11.5 million subscribers, how is this possible? Rather than focusing on just the latest 3 months, let's look at WoW player growth and loss for the past 4 years.
Player's Who Have Left World of Warcraft
Since there isn't any public information on the number of players that have left WoW, let's look at players who have been BANNED in North America and EU alone.
Here are the official Blizzard announcements since their launch:
Mar 2005: 1,000 accounts
Dec 2005: 18,000 accounts
Apr 2006: 5,400 accounts + 10,700 temp banned
May 2006: 30,000 accounts
Jun 2006: 59,000 accounts
Oct 2006: 76,000 accounts
Nov 2006: 105,000 accounts
Mar 2007: 100,000 accounts
Apr 2007: 114,000 accounts
Blizzard was banning approximately 100,000 accounts EVERY MONTH until they abruptly stopped making announcements. It's now been 3 years since they have made any official announcements regarding player banning, even though the banning still continues to happen.
Well, of course they stopped making announcements. Nothing shakes customer and investor confidence like knowing over a million accounts per year were being banned! =]
* UPDATE:
In May of 2008, it's estimated that Blizzard banned 350,000 to 500,000 accounts over a 3 day period in one of the biggest ban waves ever (most Glider accounts were associated with this ban).
Blizzard never made any official announcements regarding this ban though. It's not unheard of though for Blizzard to ban several hundreds-of-thousands of accounts. Remember when they banned 320,000 B.Net accounts in Apr 2010 and 350,000 B.Net accounts in Nov 2008?
Also of interesting note, check out the exhibits from the MDY v Blizzard lawsuit. Part of the evidence is that MDY sold over 100,000 Glider keys and that the majority of Glider users owned multiple WoW accounts (several had 3 or more). Glider users, on average, purchase one additional WoW account.
What do these trends show us?
There are, of course, many variables that leave open margin for error (e.g. abandoned accounts, players that get banned and stop playing, hacked accounts, etc.) but all of those figures are negligible and to keep things simple, the approximations below will favor Blizzard's figures.
The banning trends showed that approximately 1.0 to 1.5% of the total subscriber base were being banned on a month-by-month basis.
Even though that number was increasing, let's assume that only 1.0% were being banned per month.
On the other side of the coin, World of Warcraft's growth rate is approx. 2.0% per month (1.9138% per month to be exact) starting from the same time as the bans. (Note: Between Mar 09 and now, their growth rate is actually negative, but let's ignore that factor for now.)
So we know that almost 500,000 players were banned by April 2007. Between Oct 06 and Apr 07, World of Warcraft's user base grew from 7 to 8 million players. However, during that exact same time frame 400,000 players were banned.
The mathematics need balance.. so if WoW is growing by 2% per month (Active Subscriptions) but they are also banning 1% per month.. where does that extra 1% come from that properly balances the total number of subscribers?
When a player gets banned in the middle of the month, they typically activate a new account. However, their subscription was already paid for at the beginning of the month. On the books, that amounts to 2 "Active Subscriptions" per month because the banned account expires at the end of the month. =]
That missing 1% are actually WoW sales by banned players who purchase a new account within the same month or next.
So "active subscriptions" are actually double the number of real players. Blizzard's statement that they have 11.5 million "Active Subscriptions" is indeed accurate.. but with a twist, they're not referring to real players on those accounts. They're paid subscriptions, not players.
Summary
According to their official financials, there are currently 6.7 million paying customers. On average, about 100k players were banned per month, however since their growth rate has reduced the ban-rate would go down as well. So there could be anywhere between 6.4 million to 6.7 million real active players currently.
That number actually makes a lot more sense and satisfies their sales figures of WoW and the two expansion packs.
And according to growth (vs banned) statistics over the past 4 years, there were probably between 5.75 and 6.9 million real players during it's prime. Based on annual averages, the subscription count would be double that number, giving you the "11.5 million active subscriptions" figure. Even after favoring Blizzard for high and low figures, it's plain to see that during the course of their constant growth (see chart) about 50% of that profit was made by player-bans who purchased new accounts.
I think what had happened was that at one point World of Warcraft reached about 7 million players, but 40% of those players were banned and they purchased new accounts with new subscriptions which then created their "11.5 million Active Subscriptions" boast (averaged over several months to a year).
Blizzard has also not updated their official "11.5M" statement in over 2 years, so that is the figure that all of their employees "stick with" whether it's accurate or not today. It reminds me of Cingular's "fewest dropped calls" marketing strategy. They aggressively marketed a statistic based on a singular group's highly questionable findings and told the world that out all of all of their competitors, they had the "Fewest Dropped Calls" and the "Allover Network" citing said group as "the leading independent research company". =]
Their marketing strategies are actually quite brilliant and I admire it from a business perspective. I've discussed Blizzard's reputation management in the past and this is just another great example: It's all about wording and how they perceive the numbers. =] They've lead the entire world to believe that they have 11.5 million active subscribers (thinking "players"), when in fact it was really half that amount (possibly less than half.)
(Note: There are players with multiple accounts (ie, 2 accounts, 5-boxers, 10-boxers, etc), but it's a very small percentage and left out of these figures.)
Since growth has stopped and player-bans are down (50% of profit), this is actually a very excellent time for release of their expansion pack, it will help bolster their profit from sales and allow them to restart the cyclical process mentioned above.
So, this is just another interesting side of Blizzard most people don't know about. I love the company, not just from a gaming perspective (I've bought all of their games and will continue to do so) but also from a business perspective.. all of their business processes and methods are as polished as their games.
* Update:
Just a quick note regarding a comment made by one of the readers. Their net revenue / gross profit would also contain the sale of services and virtual items. (For example, transfers, race/faction changes, the Celestial Steed, etc.) I've been giving Blizzard favorable numbers to determine maximum possible number of players. However, if their net revenue contains 20% sales of services, for example, that means that there are even LESS players. So there could be anywhere been 6 and 7 million players, assuming that all profits are based on subscriptions. Include sales of anything else WoW related - and that reduces the number of possible players drastically.
Keep in mind, that merchandise sales are separate. The authenticator, tshirts, mugs, WoW keyboard/mouse, etc are not included in this sales figure (different vendors.)
There's a wealth of information out there on Blizzard's announcements or meetings that describe their cost of development and operating expenditures, however many of them are down-right contradictory.
Note: When I refer to Operating Costs by the way, I'm talking about EVERYTHING related to operating the WoW Servers and Employees (e.g. staffing, customer support, their Avaya phone switches & sets, HP servers, electricity, etc.)
Here are the top 5 common beliefs:
1. WoW cost $100 million for development and total upkeep for 4 years (Source)
2. WoW cost $200 million for development and total upkeep for 4 years (Source)
3. WoW cost $200 million for total upkeep but not development (Source)
4. WoW cost $100 million for development and/or total upkeep (Source)
5. WoW cost $63 million for development that took 4.5 years (Source)
If you read the articles and comments, you can certainly understand that there is a lot of confusion over this number.
Let's take the most expensive number - that WoW upkeep costs $200 million over 4 years (it's 47 months actually). On average, that's $4.25M per month.
That number seems extremely high considering that there are only ~2.5k servers (~15k blades / ~730 realms worldwide) utilized. Until that is, you read this:
"That's $200 million for the total cost of upkeep since the game's November 2004 release (presumably not including the initial cost to develop the game). This includes payroll for the entire staff, hardware support, and -- apparently the biggest infrastructure cost -- customer service."
That makes more sense: the biggest upkeep cost in that $200 million figure is customer service. So, that $4.25M per month figure is probably a lot less now considering that there have been significant numbers of lay-offs in the WoW offices and support teams, and that the server count has pretty much remained the same since the announcement.
Also, some Blizzard employees have stated that they have reduced yearly operating costs by $500,000 and have introduced new policies that have resulted in multi-millions of yearly loss reduction. (Link)
So it's safe to say that all World of Warcraft operating costs are at max $4.25 million per month. Realistically though, it's probably half of that amount or less now.
Keep this figure in mind, it's important for another post that I've been preparing. :)
It appears that Activision Blizzard's quarterly earnings report came out yesterday after my post, it actually addresses one of the updates I had provided in regards to Blizzard's Next-Gen MMO - ie, Activision is investing heavy into the new MMO, which will kick development into high gear "very soon" as previously stated.
They'll be moving employees over from other departments (e.g. Cataclysm team) within the next few months.
What's funny about the Activision Blizzard news release though, is that people think that this is "new" information. Hundreds of online news services have been posting that Blizzard's MMO will be based on a NEW IP which is apparently "new and exciting" information. That information however is actually over a year old.
So anyways, one item I forget to mention in the previous post was the release date for the game. As mentioned, even though they've been working on the "next-gen MMO" for 3 years - there hasn't been a lot of progress. Analysts have predicted that the game will be released by 2012 or 2013. However, at their current pace (although development will start increasing soon), a release date of 2015 is more likely (with beta testing in 2014).
A streaming MMO is a possibility for the next gen consoles (PS4/XBOX720) but I still have no confirmation on that from my source. There have only been some discussions and some unrelated experimentation (from the next gen MMO project). And there's basically no team tasked with creating it as of yet (not until they confirm that it can actually work well).
Other than that, Blizzard biggest concern for their product on next gen consoles is the interface.. if motion-tracking software is a bust (which it is), expect standard keyboard/mouse setups for the consoles. Blizzard will probably release customized kb/mouse hardware as well for the consoles. (Gotta milk revenue wherever you can right?)
Note: The reason I say motion-tracking tech (like Kinect) is a bust for their next-gen MMO is because it's not really applicable for twitch-based/fast-action combat. It's fine for dancing, jumping, and fun kids games.. but not a game with consequences and competition where accuracy of commands and every second is critical. As it stands, the retail product has been called very laggy, unresponsive, and inaccurate. Check out Microsoft's faked Starwars Kinect footage by the way.. the actor mimicked everything the character on the screen did (the entire exercise was probably practiced over-and-over to make it look real). It's not good when a product is constantly failing during demonstrations.